我有点失落,真的迷失了。
我试图旋转一个立方体(现在只围绕y轴),这是(丑陋和错误的)结果:
这是旋转矩阵的代码:
def rotate(axis: Vector3, angle: Float): Unit =
{
val cosAngle: Float = Math.cos(angle).toFloat
val sinAngle: Float = Math.sin(angle).toFloat
val oneMinusCosAngle: Float = 1.0f - cosAngle
val xy: Float = axis.x * axis.y
val xz: Float = axis.x * axis.z
val yz: Float = axis.y * axis.z
val xs: Float = axis.x * sinAngle
val ys: Float = axis.y * sinAngle
val zs: Float = axis.z * sinAngle
val f00: Float = axis.x * axis.x * oneMinusCosAngle + cosAngle
val f01: Float = xy * oneMinusCosAngle + zs
val f02: Float = xz * oneMinusCosAngle - ys
val f10: Float = xy * oneMinusCosAngle - zs
val f11: Float = axis.y * axis.y * oneMinusCosAngle + cosAngle
val f12: Float = yz * oneMinusCosAngle + xs
val f20: Float = xz * oneMinusCosAngle + ys
val f21: Float = yz * oneMinusCosAngle - xs
val f22: Float = axis.z * axis.z * oneMinusCosAngle + cosAngle
val t00: Float = this.m00 * f00 + this.m10 * f01 + this.m20 * f02
val t01: Float = this.m01 * f00 + this.m11 * f01 + this.m21 * f02
val t02: Float = this.m02 * f00 + this.m12 * f01 + this.m22 * f02
val t03: Float = this.m03 * f00 + this.m13 * f01 + this.m23 * f02
val t10: Float = this.m00 * f10 + this.m10 * f11 + this.m20 * f12
val t11: Float = this.m01 * f10 + this.m11 * f11 + this.m21 * f12
val t12: Float = this.m02 * f10 + this.m12 * f11 + this.m22 * f12
val t13: Float = this.m03 * f10 + this.m13 * f11 + this.m23 * f12
this.m00 = t00
this.m01 = t01
this.m02 = t02
this.m03 = t03
this.m10 = t10
this.m11 = t11
this.m12 = t12
this.m13 = t13
this.m20 = this.m00 * f20 + this.m10 * f21 + this.m20 * f22
this.m21 = this.m01 * f20 + this.m11 * f21 + this.m21 * f22
this.m22 = this.m02 * f20 + this.m12 * f21 + this.m22 * f22
this.m23 = this.m03 * f20 + this.m13 * f21 + this.m23 * f22
}
它的灵感来自:https://github.com/LWJGL/lwjgl/blob/master/src/java/org/lwjgl/util/vector/Matrix4f.java,它不再是lwjgl 3的一部分
立方体本身由这些顶点,索引和纹理坐标组成
val vertices: Array[Float] = Array(
-0.5f,0.5f,-0.5f,
-0.5f,-0.5f,-0.5f,
0.5f,-0.5f,-0.5f,
0.5f,0.5f,-0.5f,
-0.5f,0.5f,0.5f,
-0.5f,-0.5f,0.5f,
0.5f,-0.5f,0.5f,
0.5f,0.5f,0.5f,
0.5f,0.5f,-0.5f,
0.5f,-0.5f,-0.5f,
0.5f,-0.5f,0.5f,
0.5f,0.5f,0.5f,
-0.5f,0.5f,-0.5f,
-0.5f,-0.5f,-0.5f,
-0.5f,-0.5f,0.5f,
-0.5f,0.5f,0.5f,
-0.5f,0.5f,0.5f,
-0.5f,0.5f,-0.5f,
0.5f,0.5f,-0.5f,
0.5f,0.5f,0.5f,
-0.5f,-0.5f,0.5f,
-0.5f,-0.5f,-0.5f,
0.5f,-0.5f,-0.5f,
0.5f,-0.5f,0.5f
)
val indices: Array[Int] = Array(
0,1,3,
3,1,2,
4,5,7,
7,5,6,
8,9,11,
11,9,10,
12,13,15,
15,13,14,
16,17,19,
19,17,18,
20,21,23,
23,21,22
)
val textureCoords: Array[Float] = Array(
0,0,
0,1,
1,1,
1,0,
0,0,
0,1,
1,1,
1,0,
0,0,
0,1,
1,1,
1,0,
0,0,
0,1,
1,1,
1,0,
0,0,
0,1,
1,1,
1,0,
0,0,
0,1,
1,1,
1,0
)
它的模型矩阵计算如下:
def calculateModelMatrix(position: Vector3, rotation: Vector3, scale: Float): Matrix4 =
{
val matrix: Matrix4 = Matrix4.Identity
matrix.translate(position)
matrix.rotate(new Vector3(1,0,0), Math.toRadians(rotation.x).toFloat)
matrix.rotate(new Vector3(0,1,0), Math.toRadians(rotation.y).toFloat)
matrix.rotate(new Vector3(0,0,1), Math.toRadians(rotation.z).toFloat)
matrix.scale(new Vector3(scale, scale, scale))
matrix
}
从正面渲染立方体就像魅力一样。我还没有实现移动"相机",那么也许viewMatrix错了?
每帧(在相机中)计算ViewMatrix,如下所示:
def calculateViewMatrix(): Matrix4 =
{
val matrix: Matrix4 = Matrix4.Identity
matrix.rotate(new Vector3(1,0,0), Math.toRadians(this.pitch).toFloat)
matrix.rotate(new Vector3(0,1,0), Math.toRadians(this.yaw).toFloat)
matrix.translate(new Vector3(-this.position.x, -this.position.y, -this.position.z))
matrix
}
如果您需要额外的代码,我可以提供所有内容,我只是不想发布所有代码并劝阻很多人。
修改:
根据评论添加着色器代码和projectionMatrix生成:
def calculateProjectionMatrix(): Matrix4 =
{
val aspectRatio: Float = 1024 / 768 // TODO get this from somewhere
val yScale: Float = ((1.0f / Math.tan(Math.toRadians(FOV / 2f))) * aspectRatio).toFloat
val xScale: Float = yScale / aspectRatio
val frustumLength = FAR_PLANE - NEAR_PLANE
val matrix: Matrix4 = Matrix4.Zero
matrix.m00 = xScale
matrix.m11 = yScale
matrix.m22 = -((FAR_PLANE + NEAR_PLANE) / frustumLength)
matrix.m23 = -1.0f
matrix.m32 = -((2.0f * NEAR_PLANE * FAR_PLANE) / frustumLength)
matrix
}
(是的,窗口测量值匹配1024 * 768)
projectionMatrix设置一次,因为它永远不会改变。
着色码:
#version 330 core
in vec3 position;
in vec2 textureCoords;
out vec2 passTextureCoords;
uniform mat4 modelViewProjectionMatrix;
void main(void)
{
gl_Position = modelViewProjectionMatrix * vec4(position, 1.0f);
passTextureCoords = textureCoords;
}
并且,每个帧计算(并设置)modelViewProjectionMatrix,如下所示:
modelViewProjectionMatrix = Matrix4.multiply(viewProjectionMatrix, modelMatrix)
viewProjetionMatrix是:
def calculateViewProjectionMatrix(): Matrix4 =
{
Matrix4.multiply(this.projectionMatrix, this.viewMatrix)
}
并且,100%肯定,乘法 - 我们在这里有一个对象(就像所有java-devs的静态方法)
def multiply(left: Matrix4, right: Matrix4): Matrix4 =
{
val matrix: Matrix4 = new Matrix4(left)
matrix.multiply(right)
matrix
}
那里有一个拷贝构造函数,类实例的乘法方法是:
def multiply(right: Matrix4): Unit =
{
set(
this.m00 * right.m00 + this.m10 * right.m01 + this.m20 * right.m02 + this.m30 * right.m03,
this.m01 * right.m00 + this.m11 * right.m01 + this.m21 * right.m02 + this.m31 * right.m03,
this.m02 * right.m00 + this.m12 * right.m01 + this.m22 * right.m02 + this.m32 * right.m03,
this.m03 * right.m00 + this.m13 * right.m01 + this.m23 * right.m02 + this.m33 * right.m03,
this.m00 * right.m10 + this.m10 * right.m11 + this.m20 * right.m12 + this.m30 * right.m13,
this.m01 * right.m10 + this.m11 * right.m11 + this.m21 * right.m12 + this.m31 * right.m13,
this.m02 * right.m10 + this.m12 * right.m11 + this.m22 * right.m12 + this.m32 * right.m13,
this.m03 * right.m10 + this.m13 * right.m11 + this.m23 * right.m12 + this.m33 * right.m13,
this.m00 * right.m20 + this.m10 * right.m21 + this.m20 * right.m22 + this.m30 * right.m23,
this.m01 * right.m20 + this.m11 * right.m21 + this.m21 * right.m22 + this.m31 * right.m23,
this.m02 * right.m20 + this.m12 * right.m21 + this.m22 * right.m22 + this.m32 * right.m23,
this.m03 * right.m20 + this.m13 * right.m21 + this.m23 * right.m22 + this.m33 * right.m23,
this.m00 * right.m30 + this.m10 * right.m31 + this.m20 * right.m32 + this.m30 * right.m33,
this.m01 * right.m30 + this.m11 * right.m31 + this.m21 * right.m32 + this.m31 * right.m33,
this.m02 * right.m30 + this.m12 * right.m31 + this.m22 * right.m32 + this.m32 * right.m33,
this.m03 * right.m30 + this.m13 * right.m31 + this.m23 * right.m32 + this.m33 * right.m33
)
}
其中set(...)只设置Matrix4的var(iable)s ...所以第一行是m00,最后一行是m33。
def set( m00: Float, m01: Float, m02: Float, m03: Float,
m10: Float, m11: Float, m12: Float, m13: Float,
m20: Float, m21: Float, m22: Float, m23: Float,
m30: Float, m31: Float, m32: Float, m33: Float): Unit =
{
this.m00 = m00
this.m01 = m01
this.m02 = m02
this.m03 = m03
this.m10 = m10
this.m11 = m11
this.m12 = m12
this.m13 = m13
this.m20 = m20
this.m21 = m21
this.m22 = m22
this.m23 = m23
this.m30 = m30
this.m31 = m31
this.m32 = m32
this.m33 = m33
}
也许乘法是错误的(可能是,但对我来说真的很奇怪,因为我之前已经验证过)
答案 0 :(得分:4)
将import json, uuid
# models.py
from django.db import models
class ValidationCode(models.Model):
confirmation_code = models.CharField(max_length=255, unique=True)
form_data = models.CharField(max_length=255, unique=True)
create_datetime = models.DateTimeField(auto_now_add=True)
# views.py
def save_page(request):
# contains 3 fields
form = sample_tableform(request.POST)
if request.method == 'POST':
if form.is_valid():
confirmation_code = uuid.uuid4()
validation_code = ValidationCode(confirmation_code=confirmation_code, form_data=json.dumps(request.POST))
validation_code.save()
# send email with the code here
#redirect to code confirmation page
def valcode(request):
#contains confirmation_code field
sub_form = confirmform(request.POST)
if request.method == 'POST':
if sub_form.is_valid():
try:
code = ValidationCode.objects.get(confirmation_code=form.fields['confirmation_code'])
tableform = sample_tableform(json.loads(code.form_data))
if tableform.is_valid():
tableform.save()
except ValidationCode.DoesNotExist:
# return error
代码与链接代码进行比较,首先执行
Matrix.rotate
然后你做
this.m00 = t00
使用你刚刚修改的this.m00。你以类似的方式做了几次这样的事情。示例代码以相反的方式执行此操作,这意味着最终结果是不同的。交换代码块,一切都应该没问题。我希望LWJGL很快就会决定重新添加线性代数类!